Graph a real PyTorch forward pass
Runs one real forward pass and turns it into an interactive graph showing every module and tensor shape.
Loading page…
Coding / product dossier
Turns a PyTorch forward pass into a searchable graph of modules and measured tensor shapes.
Product brief
Coding agents can read your PyTorch code, but they never see what actually flows through the model. tensorViz runs one real forward pass and turns it into an interactive graph with every module and tensor shape. Claude Code, Codex and Cursor can search that graph and read the measured shapes, so they fix model code from evidence instead of guesses. Or skip the install: the Model Explorer has graphs of 560+ models from recent papers to browse and discuss. Free and open source (MIT).
Why we selected it
Gives coding agents measured tensor shapes rather than code alone—a concrete improvement for model debugging. MIT licensing and a browser-accessible explorer with 560+ model graphs make it accessible to researchers and/
Product preview saved with our daily selection
Capability scan
Only capabilities supported by the product information we collected are listed here.
Runs one real forward pass and turns it into an interactive graph showing every module and tensor shape.
Lets Claude Code, Codex and Cursor search the graph and read measured tensor shapes to inform model-code fixes.
Provides a Model Explorer with graphs of more than 560 models from recent papers, available to browse without installing tensorViz.
Best-fit use cases
FAQ
It turns one real forward pass into an interactive graph containing every module and tensor shape.
Claude Code, Codex and Cursor can search the graph and read its measured tensor shapes.
Yes. The Model Explorer offers graphs of more than 560 models from recent papers to browse and discuss without an installation.
Yes. tensorViz is free and open source under the MIT license.